PROJECT SUMMARY Psychiatric disorders are highly polygenic, exhibit a complex pattern of genetic correlations across the full spectrum of diagnostic categories. Genetic risk for psychiatric disorders acts via a poorly understood set of intermediate mechanisms. With substantial investments in large consortia, registry-based efforts, and national biobanks, genome-wide association studies (GWAS) of psychiatric disorders and related quantitative phenotypes have made substantial strides in attaining the power needed to detect reproducible genetic associations, estimate genome-wide chip heritabilities, and estimate genetic correlations between traits. Combined with bioinformatic approaches, GWAS efforts have produced insights into tissues and cell types relevant to psychiatric disease, and atlases of genetic correlations have rapidly expanded the ontological network of descriptive knowledge of shared genetic architecture across psychiatric diseases and social, behavioral, and biological traits. In order to more fully capitalize on this growing corpus of GWAS research output, we have recently introduced Genomic Structural Equating Modeling (Genomic SEM; Grotzinger et al., 2019; Nature Human Behaviour), an analytic framework and associated software for multivariate modelling of genetic architecture using GWAS summary data from samples of varying or unknown degrees of overlap. The primary goal of this R01 proposal is to capitalize on and further develop Genomic SEM to formally investigate genetic risk sharing across psychiatric disorders and- equally importantly- genetic differentiation between them. We will (1) identify transdiagnostic dimensions of genetic sharing across psychiatric disorders, and test for commonalities and divergence in genetic associations with biological and psychosocial dimensions of potentially cross-cutting genetic risk; (2) Identify gene sets and categories that contribute disproportionately to risk sharing across disorders and/or to disorder-specific genetic variation; (3) Formally distinguish disorder-general from disorder-specific Loci; and (4) Considerably expand the suite of methods currently available in Genomic SEM software to meet increasing demand by the genetics community. The availability of sex-stratified GWAS summary data will allow us to examine convergent and divergent patterns of association and multivariate genetic architecture across males and females. Moreover, we will incorporate cutting edge methods for modeling trans-ethnic data, which will be of increasing value as more diverse GWAS samples become available. This project will constitute the most comprehensive interrogation of the shared and disorder-specific genetic architecture of major psychiatric disorders and their relationships to biological and psychosocial dimensions of potentially cross-cutting genetic risk, and will provide an expanded suite of novel, user friendly, free, open-source tools that serve the entire genetics community.